Google on Tuesday launched Gemini Enterprise for Financial Services and a similar version for the legal industry.

The releases mark Google’s belated move toward vertical and domain-specific AI agents, following the example set by Anthropic and OpenAI, the company’s two main AI rivals.

Both Gemini Enterprise for Financial Services and Gemini Enterprise for Legal are built on the Gemini Enterprise platform. The financial package includes a Google-managed financial research agent, more than 50 new skills with specialized agentic instructions for financial roles and workflows, and enterprise data connectors. Gemini Enterprise for Legal includes specialized skills for lawyers and connectors to specialized legal systems.

Google is following a trend that frontier model creators such as Anthropic and OpenAI began more than a year ago, when Anthropic released Claude for Financial Services. Anthropic Claude Cowork also has specific plugins for the financial and legal sectors. OpenAI also offers plugins on its Codex platform used across industries, including finance. Finance and legal are among the most demanding markets for AI vendors: both are heavily regulated industries, and firms in them must satisfy compliance, data-governance and auditability requirements before AI agents can be deployed in production workflows.

Related: Prompt: Agentic AI Is Outpacing Enterprise Readiness

With other vendors already moving toward domain-specific offerings, Google appears to be launching Gemini Enterprise for Financial Services somewhat late.

“Right now, Google is playing catch-up in a lot of ways,” said David Nicholson, an analyst at Futurum Group. “However, they are trying to make up for that by focusing on openness and the idea that you’re not going to ditch all of your existing processes.”

The Question of Neutrality

Nicholson said Google is trying to appeal to enterprises by presenting itself as a neutral platform layer that works with existing processes. In a statement, Google Cloud CEO Thomas Kurian said financial professionals want a platform that does not lock them into a single model or ecosystem and that is highly secure and compliant.

However, Nicholson said enterprises may care less about lock-in than about trust. In regulated sectors such as banking and law, that trust has a practical dimension: firms answer to regulators for how AI is used in their operations, which makes guardrails, security and auditability central to purchasing decisions.

“[Enterprises] are seeking to trust that agents work within frameworks and guardrails that a company trusts,” Nicholson said. He added that enterprises already within Google’s ecosystem know there are defined guardrails they can operate within.

“They are more likely to trust Google on this than they are Anthropic or OpenAI as an individual standalone company,” Nicholson continued.

Weaknesses and Strengths

Google’s challenge, Nicholson said, is that it has fallen somewhat behind competitors such as OpenAI and Anthropic. Although Gemini was at times ahead of ChatGPT and Claude in performance, it has since been overtaken, leaving Google with significant ground to make up.

Related: Nvidia’s New Open Model Is for Specific Use Cases

Nicholson also argued that Google’s neutral-platform pitch is weakened by its role as a major participant in the AI race. Because it has an incentive to stay ahead of OpenAI and Anthropic, he said, it cannot be fully neutral.

“It’s unclear how much Google is going to gain from this argument that they’re neutral,” he said, adding that vendors less tied to the AI competition could be more attractive to enterprises that do not want to be dependent on a single provider.

Google’s stronger position may instead come from its ability to give financial services firms access to applications that support their workflows, often built on systems from CoinDesk Data & Indices, Daloopa, Dun & Bradstreet and FactSet, according to Sid Nag, founder and chief research officer of Tekonyx.

“The strategic asset is not Gemini by itself, but Google’s ability to connect Gemini Enterprise through the connectors that are talked about to licensed data repositories in the area of financial industries,” Nag said.

He added that Google could have an advantage by offering a “strong enterprise control plane.”

Even so, Nag said the new Gemini Enterprise financial and legal systems show that “Google is correctly shifting enterprise AI conversation from model capability to industry work for execution.”

For buyers, the launch presents a concrete test case: whether domain-specific packaging and connectors to licensed data can outweigh reservations about relying on a vendor that is itself a major competitor in the AI race.

Related: Meta Reverses Course with Open-Weight Muse Glimmer